Healthcare providers of severely neurological ill and critical patients are faced with the difficult task of communicating neurological prognosis to the families of patients who are often incapacitated or comatose during the acute phase. Withdrawal of life-sustaining treatments (WLST) accounts for more than a third of deaths in neurocritical care, and there is a lot of heterogeneity in the data about possibly avoidable mortality. Guidelines are now available for neuroprognostication in the survivors of cardiac arrest, severe acute brain injury (SABI), spinal cord injury (SCI), and Guillain–Barré syndrome (GBS), but the current evidence is mostly of moderate-to-low quality. Certain well-validated prognostic models, rather than individual clinical variables, seem to be moderately reliable in predicting defined outcomes. There is a lot of scope for the improvement of the quality of evidence and the formulation of decision aids for clinicians to communicate neurological prognosis with more certainty and enable surrogates to make informed decisions, along with healthcare teams.